Psychiatrists lack markers that would allow them to select treatments for patients with major depressive disorder (MDD) based on the likelihood of response in individual patients. Identifying robust predictors of treatment outcome could greatly reduce the morbidity and mortality resulting from ineffective treatment trials. Cognitive-behavioral therapy (CBT) is a manualized treatment with well-established effectiveness for MDD. Converging evidence suggests that in vivo quantification of the serotonin 1A (5-HT1A) receptor in two specific brain regions, the orbitofrontal cortex and raphe nuclei, may be predictive of outcome with CBT for depression. We and others have found higher 5-HT1A receptor binding in these regions to be associated with worse treatment outcome with either naturalistic treatment or medication-based treatment for depression. Pharmacogenetic studies have found that a functional polymorphism in the 5-HT1A receptor which contributes to elevated 5-HT1A binding in the raphe nuclei is associated with diminished antidepressant response. 5-HT1A receptor binding in the raphe nuclei is inversely associated with amygdala reactivity to negative emotional stimuli, which itself has been found to be predictive of outcome with CBT for depression. Lastly, 5-HT1A receptors modulate learning processes likely to be recruited by CBT for depression, with evidence of regionally specific effects in the prefrontal cortex. We will quantify the 5-HT1A receptor at baseline using positron emission tomography (PET) with the [11C]WAY-100635 radioligand among 46 subjects with MDD in a major depressive episode, and to administer standardized individual CBT for MDD over 12 weeks, assessing baseline 5-HT1A binding potential in the orbitofrontal cortex and raphe nuclei as predictors of outcome. We hypothesize that elevated 5-HT1A binding potential in these regions will predict worse outcome with CBT for MDD. Reversal learning is a process in which subjects must establish and then reverse associations of specific stimuli with either punishment or reward, and is considered a measure of cognitive flexibility. Less quantitative measures of cognitive flexibility have been associated with better outcome with CBT for depression. We will assess performance on a probabilistic reversal learning task at baseline among subjects undergoing subsequent CBT for MDD, hypothesizing that better task performance will be predictive of better treatment outcome. Finally, there is strong evidence that serotonin acting in the prefrontal cortex modulates reversal learning performance, and evidence from animal studies that the 5-HT1A receptor plays an inhibitory role in reversal learning performance. We will therefore assess the degree to which 5-HT1A binding in the orbitofrontal cortex is correlated with reversal learning performance in humans, to clarify the neurochemical modulators of this type of learning. These studies may suggest novel, rational combinations of pharmacotherapy and psychotherapy for depression based on an increased understanding of the neurochemical modulators of learning relevant to CBT for depression. The Candidate Dr. Miller studied medicine at Yale University, and conducted residency training in psychiatry as well as post-doctoral research training at Columbia University. His short-term career goals are to become expert in PET imaging methodology and analysis in order to apply this technique to study the pathophysiology of depression, and to the mechanisms of existing and potential new treatments. Another major goal is to gain expertise in conducting rigorous research of psychotherapy in a translational manner. Long term goals include the development of studies in which subjects are randomized to multiple treatment conditions with assessment of baseline predictors that may predict differential response to specific treatments, as well as the development of more effective antidepressant treatments based on an increased understanding of underlying pathophysiology. Environment The New York State Psychiatric Institute and Columbia University's Department of Psychiatry have outstanding institutional resources for the development of Dr. Miller's career, from the intellectual expertise in PET imaging research by his mentor and co-mentor (Drs. Parsey and Mann), to the physical resources available to him, including a 3T MR scanner; three PET scanners in the Kreitchman PET Center used for human PET imaging; a computer cluster to support image analysis; the Radioligand Production Laboratory directed by Dr. Dileep Kumar; and office space in The New York State Psychiatric Institute.